The sociology of poker tells
I recently read Cate Hall's 2024 article about agency. It's probably the first article I ever found by pre-ordering the book that grew out of that very article.
Cate has a poker background and notes that she got a big edge from studying physical reads. This is not surprising to me at all: I was always much more tell-oriented1 than most of my peers, and I got a lot of eye-rolls for saying so. A few notes:
- Emphasizing physical information violates a lot of strong players' ideas not just of how to study effectively but of what poker is. To them, caring a lot about physical reads is abandoning the real thing for a cartoon.
- Similarly, and closely relatedly: I've maintained for many years now that newer generations of poker players are not nearly aware enough of cheating. Again, it offends many of these players' sense of what poker is.
- Paying too little attention to physical reads is also a great example of over-emphasizing quantitative kinds of analysis. Of course, when you get a read, you should incorporate it into a quantative decision-making process.2 But there is a comforting certainty to studying a solver output that isn't there when you're studying the physical side of the game.
- Getting physical information requires a certain receptiveness to the people you're sharing a poker table with. Or, at least, it requires paying a lot of attention to them. Many people don't want to do this, often because they find it boring or they simply don't like their average table-mate very much.
- Many players, especially weaker ones, give off more physical information when the stakes get higher or decisions get more important. (I found the Main Event to be an absolute fiesta of tells.) Players who, implicitly or explicitly, judge the value of physical information only by the frequency with which they find it are likely to be making an expected value mistake.
I don't really think of it as "tell-spotting," because the term suggests discrete, Oreo-cookie-type behaviors, and a lot of physical information isn't like that. I thought of physical information as being "more in the adverbs than the nouns."↩
This often happens in a probabilistic way, or around the edges of a hand range. ("Would he really look like that with aces? Nah." "Eh, this feels a bit more like a draw to me.")↩